ESP32-C3 power consumption when held in reset

[peteswenson]

On a custom board, I’m using an ESP32-C3FN4 (chip down) running AT firmware v2.4 to scan for nearby WAPs. From the main MCU, I control the ESP32 with its EN pin. For testing, I have a pushbutton on EN to Ground. I’m using a PPK2 to measure current.

After reset and before any configuration, I get ~25mA. After I send AT+GSLP=3600000, I get 5-10mA. When I hold the ESP32 in reset, I get 11-12mA. I would expect the power consumption to be the absolute lowest in reset. Could this be due to running the AT firmware? Are there other AT commands I could send to help?

[ESP_Sprite]

The chip is entirely off when held in reset; the exact power use is in the datasheet but iirc should be below an uA. Are you sure you do not have something else leaking power, e.g. pullups/downs?

[peteswenson]

I'm pretty sure, but I'll keep looking...
Thanks.

[peteswenson]

I cut the power traces to all of the peripherals: main MCU, external memory chip, accelerometer, and battery charger. I also removed the 3.3V LDO, shorted the input pin to the output pin, and set the PPK2 output to 3.3V. I also cut the EN trace between the main MCU and the ESP32. There's a pulldown on EN so its default is Off. Here are the current measurements:

After reset, avg = 20.58mA
After AT+SLEEP=2, avg = 1.09mA
Held in reset, avg = 4.75mA
Pull Tag-Connect cable (no terminal interface), avg = 5.19mA

There's nothing left that explains the 4-5mA when in reset, unless the ESP32 is doing stuff. So I erased the ESP32:

esptool.py --port COM5 --baud 115200 erase_flash

After erasing the ESP32, avg = 8.81mA
Then held in reset, avg = 5.79mA

[peteswenson]

Any ideas?

[ESP_Sprite]

I don't know what to tell you... something is still leaking current somewhere on your design. Might be as simple as a resistive divider, might be something as flakey as slightly-conductive flux residues. For instance, you don't happen to have a pullup on EN that conducts current to ground when you pull EN down?

[ESP_Sprite]

For reference: this is an ESP32-C3-Wroom-02 (containing the same ESP32-C3 chip) when held in reset. I measure the fraction of an uA that I'd expect.

esp_reset.jpeg (418.34 KiB) Viewed 2610 times